AWS’ mission to evenly distribute the future
Amazon Web Services’ message has been a consistent one over the last few years I’ve been coming to its AWS Summit events: “come on in, the infrastructure’s lovely” (i.e. why are you not with us already?). Thumping dance music intros? Check. Senior Exec flown in to evangelise? Check. Parade of customers given their six minutes each of stage-time to illustrate the advantages inherent in the Way of the Cloud (in amongst the product launch plugs and “we’re hiring” promos)? Check.
Nuances have changed along the way, of course. Graphics depicting cost reductions, headache-free heavy-lifting (upon which you’re free to add your own value), faster time to production, and the freedom to fail fast and cheaply as an enabler of innovation (whether in start-up disruptor, or agile enterprise team) were all present and correct in the slide-deck. Such points have been joined more recently by AWS promoting its hybrid credentials to help customers journey to the cloud, and big enterprise names declaring their love for being ‘all in’ on AWS (having completed that migration journey).
So what’s this year’s focus? In part, more of the same – for example:
- FanDuel declared that whilst it considers its infrastructure team to number in the 100s, only 12 people actually work for the company – the rest are resources it feels it can call upon from AWS. Its fantasy sports gaming platform needs to scale from 10s of transactions a minute to 1,000s every second at peak periods (currently, during Sundays’ NFL games in the US), and an AWS elastic infrastructure helps it do that.
- Travelex is busy transforming itself from a bricks-and-mortar travel money shop to a multi-channel digital business for global travellers’ financial services. Applications for Its new SuperCard credit card broke the registration website on its old provider, but when prospective ‘traditional developers’ were quoting months of dev time and £100k’s of cost to fix the problem, it decided to build a solution itself to handle the load using various AWS services. Four weeks later, the effort of just two developers has produced what the company calls ‘SuperQueue’ – an ‘infinite buffer’ which has been successfully stress-tested to cope with 400,000 new registrations an hour, and has so far cost less than £1,000 to run.
- Trainline has managed 10,000 releases the last three months; pre-AWS it managed just one every six weeks.
- The UK’s Driver and Vehicle Standards Agency has rolled out a replacement for its MOT system using agile methods which has opened up access to testing data to mobile and through APIs to third-parties, at half the cost of the somewhat more closed and static system in place over the last ten years.
But if I’m saying I’ve heard these stories (or similar) before… what here is news in Cloud Land for 2016, and what’s differentiating AWS?
Amazon.com’s CTO Werner Vogels described how Amazon itself broke up its initial monolithic application of the late 1990s into services for customers, products and orders; and further into microservices more recently as it became clear the unit for scalability needed finer granularity to offer greater flexibility and efficiency as the company grew.
From AWS’ earliest days of generic storage and compute services, the company has expanded its roster to cater for more and more workloads – specific to databases, security, application deployment, IoT, etc. – as it moves its ‘heavy-lifting’ capabilities further up the stack (and allows its customers to consign more and more of their former infrastructure headaches to a third-party’s services on PAYG contracts).
Where’s the new focus? There’s sessions on the company’s IoT platform. It’s not the only IoT platform in town, of course (and IBM’s partnership with Cisco makes much of their ability to sink some compute power at the edge, rather than relying on shipping all your data back to Cloud Central for processing). But AWS’ strength is in its scale and the breadth of services it can offer up to customers prepared to put in the effort to glue them together (or pay partners to do it for them). So, IoT-friendly messaging, an IoT Rules Engine, Device SDKs, all coupled with access to services like AWS Kenesis, Lamda, S3, etc to analyse and store the data, and build apps on top that do something with it.
Interestingly, although AWS does boast machine learning services, unlike other cloud providers it’s chosen not to dwell whether / how AI will change the world. Maybe it doesn’t believe that hype (yet), maybe its grounding its promotional activity more in the here-and-now world of the 1,000s of developers here eager to hear how to build tomorrow’s app first (rather than next week’s). Or maybe we’re all ‘post-cognitive’ already.
In short, AWS’ message is that it has not only the building blocks, but also the machinery to lift them for you, so you can transform your business into the agile, mobile world of cloud with the minimum of risk and fuss. It’s striving to make it no-brainer easy to make a move to the cloud, and if you wait for the bar to drop any lower the chances are you’ll have already been disrupted out of your market by someone who got on board earlier.
William Gibson might not have though that the future was evenly distributed yet, but AWS seem fervently intent on distributing it far and wide so everyone can be a customer. After all, you don’t build a $10bn annual run-rate company in ten years by shying away from the mass market.